USP Theses Collection


Menu



A B C D E F G H I J-K L M N O P Q-R S T U-V W-Y
close this section of the library Vehicles, Remotely piloted


View the PDF document A swarm model for planar formations of multiple autonomous unmanned aerial vehicles (UAVS)
Author: Sharan, Ashna
Institution: University of the South Pacific.
Award: M.Sc.
Subject: Drone aircraft -- Control systems, Drone aircraft -- Automatic control, Drone aircraft -- Mathematical models, Vehicles, remotely piloted
Date: 2013
Call No.: pac UG 1242 .D7 S53 2013
BRN: 1191920
Copyright:40-60% of this thesis may be copied without the authors written permission

Abstract: This thesis addresses the common ndpath problem for a multi-agent robotic sys- tem, the solution of which has been inspired from observations of multi-agent dynamical systems in nature including certain species of insects, birds and sh. For this reason the introductory chapter is devoted to understanding relevant con- cepts of swarming in the context of biological systems in nature which then serve as the building blocks of the emerging eld of swarm robotics. The primary ob- jective is to control multiple autonomous quadrotor Unmanned Aerial Vehicles (UAVs) simultaneously. In order to meet this objective, a general swarm model for a system of simple point-like rigid bodies is designed. The general swarm model is then adapted so that it is applicable to a system of quadrotor UAVs. The Direct Method of Lyapunov is used in designing control algorithms to ensure the stabil- ity of the entire system. Lyapunov functions are constructed and new velocity controllers are derived as feedback controllers via consideration of their gradients. Two cases are considered, in the rst case the conguration space is free of static obstacles and in the second it is cluttered with static obstacles. Once the velocity controllers have been dened, the swarm model is simulated for verication of its functionality. Basic patterns of formation which are similar to emergent behaviors distinctive of swarming dynamical systems in nature are demonstrated. Some ad- vanced maneuvering such as split and rejoin and tunnelling are also observed in the stationary obstacle collision avoidance scenario. Dierent emergent pat- terns are obtained with variation in the control parameters. Time evolution of the translational and angular velocities derived are also evaluated to understand the emergent patterns exhibited. v
View the PDF document Development of an automatic guided vehicle with an obstacle avoidance system
Author: Kumar, Shivendra.
Institution: University of the South Pacific.
Subject: Robots -- Control systems , Vehicles, Remotely piloted , Intelligent control systems
Date: 2003.
Call No.: pac TJ 211 .35 .K85 2003
BRN: 1055845
Copyright:Under 10% of this thesis may be copied without the authors written permission

Abstract: Optical guidepath following automatic guided vehicles (AGV) have its applications limited due to its inability to avoid obstacles that block its path forward. This research therefore develops an AGV that follows optical guidepath for navigation and is able to able to maneuver around obstacles. An algorithm that detects obstacles and avoids them is presented first. This algorithm, called the guidepath following certainty grid (GFCG) method uses data from infra red (IR) ranging sensors located in front of the AGV to detect and avoid obstacles. After maneuvering around the obstacle using the contour of the obstacle, the AGV rejoins the guidepath and resumes its movement towards its destination. Details on the framework of the AGV, its control circuitry and organization of intelligence are then discussed. The aluminum framework housing the circuitry and load/ unload unit uses DC motors for movement. The AGV receives information about its environment from IR sensors in the front and the side, line tracing sensors and a hall effect compass. Coordination and processing of intelligence is performed by a PIC16F877 microcontroller. Experimental results on line tracing, response distances and obstacle avoidance are presented thereafter. Results in line tracing and response distances exhibit orientation errors of the AGV. The response distance of the AGV upon detecting an obstacle or a station is also presented along with path maps taken by the AGV to avoid obstacles placed at different positions. The conclusion summarizes the features and capabilities of the AGV and discusses the implications of this research to the region. Areas of further research are also discussed.
Disclaimer & Copyright  l  Contact Us  l  
© Copyright 2004 - 2009. All Rights Reserved.
USP Library
The University of the South Pacific
Laucala Campus, Suva, Fiji
Tel: +679 323 1000